Threat Assessment of Child
Sexual Exploitation and Abuse
June 2013
Child Exploitation and Online Protection Centre
© CEOP copyright 2013
2
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Contents
Introduction
5
Key threats & General picture
6
Key threat 1: The proliferation of indecent images of children (IIOC)
8
Summary
9
Key threat 2: Online child sexual exploitation (OCSE)
10
Summary
13
Key threat 3: Transnational child sexual abuse
14
Summary
17
Key threat 4: Contact child sexual abuse
18
Summary
21
Appendix A
22
3
4
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Introduction
Aim & scope
1.
Last year the Child Exploitation and Online Protection (CEOP) Centre published its first annual Threat Assessment of Child
Sexual Exploitation and Abuse (TACSEA), which examined current and emerging threats posed to children in the United
Kingdom from sexual exploitation and sexual abuse.
2.
The 2013 TACSEA builds on and updates last year’s document and, in line with CEOP’s founding ethos that every child
matters…everywhere, also examines the threat to children abroad from UK nationals. The purpose of the assessment is to
enable CEOP to set its strategic priorities for the year ahead by describing the nature and extent of the threat landscape
and how we assess it will change during the year. This will ensure that in a time of austerity across the public sector, CEOP
resources are deployed where the threat to children is the greatest. It also provides a picture for partners to consider in their
strategies and resource deployment.
3.
CEOP’s mission, as set out in its three year strategy, is to work with others to protect children from sexual abuse and
exploitation. It is recognised, however, that many other agencies in the UK have responsibility for tackling the spectrum of
child sexual abuse particularly where it occurs in the home or within the family. Whilst acknowledging the wider context of child
sexual exploitation and abuse this assessment focuses on those areas, both online and offline, where CEOP can add value
and where no other agency has an appreciable footprint.
4.
The TACSEA is an intelligence product, the purpose of which is to enhance understanding of a particular threat area. As an
exception reporting tool it highlights only significant changes to that threat during the reporting period and it should therefore
be read in conjunction with the 2012 TACSEA. The assessment sets out what is currently known about the threats to children
and highlights those areas where understanding is less well developed.
5.
This assessment signals the start of the second year of a three year strategy in which CEOP undertakes to assess where and
how children are most at risk from sexual exploitation and abuse, to communicate this widely and to develop programmes to
mitigate threats. It is likely that CEOP will conduct an additional threat assessment in the autumn of 2013 to inform the setting
of national priorities.
6.
In October 2013, CEOP will become a command of the National Crime Agency (NCA). The NCA will be the centrepiece of the
reformed policing landscape and will spearhead the national crime fighting response to serious, organised and complex crime.
Within the NCA, CEOP will further develop its strategic approach to prevent child sexual exploitation and abuse, to protect
children and young people who are at risk of victimisation and to pursue offenders who target children in the UK or overseas.
Information sources and limitations
7.
Perhaps to a greater extent than other criminality, most areas of child sexual exploitation and abuse suffer from chronic underreporting by victims and often inconsistent recording in the criminal justice process. Accurately identifying the size of the UK
offending and victim populations is therefore problematic.
8.
The analysis within this assessment is based on a combination of statistical measures, research and the professional
judgement of subject area specialists within CEOP and across the wider child protection community. As with any strategic
intelligence product the data gathered is subject to a process of analysis involving interpretation, reasoning and speculative
consideration of trends, future developments, risks and opportunities.
5
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threats & General picture
Key threats
9.
This assessment builds on the thinking in the 2012 TACSEA by identifying the high level key threats to children from sexual
exploitation and abuse. These form the basis of the CEOP Control Strategy for 2013/14. The TACSEA will be underpinned by
a strategic action plan for each key threat.
10. In the majority of its forms, both online and offline, child sexual exploitation and abuse (CSEA) remains a largely solitary crime,
albeit one often exacerbated by the effects of group dynamics. Often where true group offending does occur, this shares few
of the characteristics traditionally associated with organised crime. CSEA offending is, however, often extremely serious and
complex in both its execution and impact. The key threats identified in this document represent the most serious and complex
areas of CSEA, including those in which organised crime is sometimes a feature.
General picture
11. Recent research conducted by the NSPCC 1 indicates that around 5% of UK children suffer contact sexual abuse at some
point during childhood. It is likely that around 190,000 of these will fall victim to contact sexual abuse by a stranger or an adult
relative (other than a parent or guardian) before turning 18. This represents an average of more than 10,000 new victims in the
UK every year.
12. In addition, CEOP receives reports from around 1,000 children each year concerning online victimisation by adults. A further
unquantifiable number of children overseas suffer contact sexual abuse at the hands of UK nationals visiting or living and
working abroad.
13. In the eyes of the general public, from a child protection perspective, the second half of 2012 was dominated by reports of
historic sexual abuse, most notably the prolific and predatory sexual offending of Jimmy Savile and others in the public eye.
The scale and nature of the abuse able to be perpetrated by Savile - to a degree hiding in plain sight - were truly shocking and
abhorrent. The extent to which his offending is indicative of today’s threat landscape is, however, unclear. The abuse of the
power and authority that comes with status, be that celebrity or otherwise, certainly remains a potential threat to children. The
vulnerability of certain institutions to predatory child sexual offenders should also not be underestimated.
14. The internet and wider information and communications technology are now firmly embedded within the everyday lives of UK
children. Online activity features so prominently in their entertainment, education and social lives that it is seen by them as a
tool in their offline lives. In turn, UK children are now more accessible to offenders online, with 91% living in a household with
access to the internet. The numbers of hours 12 to 15 year olds spend online each week has risen from 14.9 hours in 2011 to
17.1 hours in 2012, an increase of almost 15%.2
15. The internet also offers children the opportunity for a separate identity in which they can be whoever they wish and take risks
they would never countenance offline. It is often where these two worlds meet that children suffer most at the hands of those
wishing to do them harm. That a child’s internet presence is now so inextricably linked to their real world identity can only
increase the impact of any exploitation they suffer.
16. Whilst the edges are blurred for children, it is important to recognise that a distinction between online and offline may remain
for adult sex offenders who offend against children. Not only are some offenders generationally less exposed to the internet but
the existence of a delineated and seemingly anonymous online environment can bring many advantages to them.
6
1
Child abuse and neglect in the UK today (Radford et al, 2011).
2
Children and Parents: Media Use and Attitudes Report, Ofcom, 2012.
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threats & General picture
17. In 2012 just over one third of the world’s population was connected to the internet. Significant regional variations in internet
penetration exist, however, from 73.2% in North America, Europe and Australasia to less than 20% in Africa. Given the
enormous populations of some countries in the developing world, the scope for growth in internet adoption is clearly huge.
The gap between the developed and developing world is likely to narrow considerably over the next few years. The threat to
children globally from sexual exploitation online can therefore only increase.
18. Whether acting alone or in groups, the primary or sole motivating factor for the great majority of CSEA offenders is sexual
gratification. Some act on a specific sexual interest in children whilst others target the vulnerability of youth more broadly for
sexual purposes. A much smaller number of threat actors sexually exploit children for financial or other gain.
7
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 1: The proliferation of indecent images of children (IIOC)
19. Indecent images of children (IIOC) continued to proliferate across the internet with no single means of storage or distribution
achieving overall dominance in 2012. With the possible exception of the live video streaming of child sexual abuse, the
distribution methodologies remain much as they were in 2011. A rise in overall levels of hidden internet use in the UK may
indicate that displaced offenders are taking advantage of this technology. Although it is clearly not possible to establish a
precise figure, CEOP estimates that there were around 50,000 individuals in the UK involved in downloading and sharing IIOC
during 2012.
Webmail, social networking and file hosting
20. Reporting to CEOP allows for an estimate as to the level of IIOC transactions taking place over webmail, social networking
sites and to a more limited extent, mainstream file hosting services. In 2012, CEOP received reports of just over 8,000 IIOC
transactions involving these facilities and a UK-based user. These reports contained a total of 70,000 still and moving IIOC
- a two-fold increase on previous year’s values. It should be noted that these are not all unique images as there is substantial
duplication in IIOC across transactions. In line with figures from 2011, a fifth of the total was assessed as having been
self-generated by their subject.
File sharing (peer to peer)
21. File sharing using decentralised networks (known as Peer to Peer or P2P) remained a popular means by which UK offenders
downloaded and distributed IIOC during 2012. P2P offenders are often more prolific in their downloading habits with vast IIOC
collections being shared or built in a matter of hours.
The hidden internet
22. The use of the hidden internet by IIOC offenders remained a key threat during 2012 with the number of UK daily users
connecting to it increasing by two-thirds during the year. This represents one of the largest annual increases globally, in a
non-oppressive regime 3. Given the rise in overall use it is reasonable to conclude that use by UK offenders as an IIOC sharing
environment increased during 2012.
The commercial production and distribution of IIOC
23. The commercial distribution of IIOC on the open internet is estimated to account for a very small percentage of the
transactions taking place. This low level is likely to be a result of the large volume of IIOC in free circulation, particularly over
P2P, and widespread awareness of the traceability of conventional payment methods. No evidence was found to suggest that
a substantial commercial market exists in 2012, although there are some indications of its slight re-emergence in certain areas
of the hidden internet and live streaming activity.
Live video streaming
24. Live streaming was identified as an emerging method of IIOC production and distribution in 2012. Offenders have been seen to
target vulnerable families overseas to facilitate live access to children over webcam. The children are made to engage in sexual
activity in exchange for payment to the family or to an organised crime group.
Image analysis
25. Little is currently known about the demographic make-up of IIOC victims and what this tells us about the threat from IIOC
offenders as few comprehensive studies exist. A sample of 118 IIOC reports received by CEOP in October 2012 revealed
a total of 146 child subjects on which to base a victimology analysis. These reports related to IIOC shared by UK based
offenders.
3
8
Based on the rankings in the Freedom House, Freedom in the World Report 2012.
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 1: The proliferation of indecent images of children (IIOC)
26. The following findings emerged:
120 of the victims depicted were female and 26 male
96% of victims were white 4 and 4% were of Asian appearance
75% of images were at level 1 on the Sentencing Council (SC) scale 5, 2% at level 2, 8% at level 3 and 15% at level 4.
There were no level 5 images present in the sample
20% of images, depicted sexual contact between a child and an adult. All of these adults were white, 76% were male and
24% were female.
27. A second analysis of IIOC was undertaken over a three year period covering January 2010 to December 2012. The aim was to establish
trends in image type and victim demographics. A sample of 321 IIOC 6 evenly distributed over the three year period revealed the following:
An overall increase in the number of female children in images
A 70% increase in female victims aged under ten
A 25% increase in female victims aged over ten
A 68% reduction in male victims under the age of ten
A 48% reduction in male victims over ten years
White victims represented 91% of the total
A 125% increase in the number of images at SC level 4 although level 1 images remain the most prevalent
A 23% increase in the number of images in which both adults and children are present
An increase in the number of female offenders in images.
28. Given the global nature of IIOC sharing, there is no reason to suspect that the image samples used in these analyses are
unrepresentative of wider IIOC in circulation by webmail and social networking. With this in mind, the under-representation of
non-white victims in the sample is of note. However, the UK locus of the individual sharing transactions reported may skew the
sample to an extent as this may have some bearing on the particular preferences of those offenders involved. Further research
is required to establish the extent of any correlation between victim and offender ethnicity.
29. The second analysis identified an apparent shift from male victims to younger female victims. Both analyses revealed a small
increase in the number of female adults involved in the sexual abuse depicted than anticipated, but this is in keeping with the
trend identified in last year’s TACSEA.
Summary
30. The overall threat from the proliferation of IIOC has remained relatively stable in 2012 in terms of overall offender numbers.
There is, however, evidence of a migration away from more open methods of distribution to the hidden internet.
31. Whilst those who view IIOC online pose a risk to any children to whom they have access, the greatest composite risk from the
proliferation of such content is assessed as coming from those involved in the production of level 4 and 5 images. Although
research suggests that those who possess images of a higher Sentencing Council classification do not necessarily pose a
higher risk to children, the production of such images will inevitably involve more serious contact sexual offending.
32. The use of the hidden internet in the UK and beyond is expected to continue increasing throughout 2013, possibly reaching
20,000 daily UK users by the end of the year. CEOP assesses that the networking and sense of ‘safe’ community that occurs
within the hidden internet and the relative sophistication of offenders within that environment stimulates the production of IIOC
on both a commercial and non-commercial basis.
33. CEOP assesses that the live streaming of sexual abuse - particularly from the developing world carries a high composite risk
in 2013. The threat results from a juxtaposition of factors which includes extreme poverty, increasing penetration of high speed
internet and poor local child protection response. In addition, the existence of a vast and comparatively wealthy overseas client
base has inevitably led to the involvement of organised criminal groups in the host nations.
4
5
6
Includes those of Latin American or southern European appearance.
See Appendix A.
Some images contained more than one child victim.
9
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 2: Online child sexual exploitation (OCSE)
34. The term online child sexual exploitation (OCSE) is used to describe a genre of internet offending which includes, but is
not defined by, traditional notions of online grooming. In this context, OCSE includes the much broader threat from online
communication between an adult and a child for the purposes of sexual exploitation. Such conduct is sometimes referred to
by the generic term of online ‘grooming’ which, whilst certainly an element of OCSE, does not necessarily define it. The threat
from grooming is considerably narrower than that sought to be assessed under this theme.
35. Use of the term ‘grooming’ also suggests a course of conduct evolving over a period of time while the offender subtly gains
the trust of his victim. Whilst slow-time grooming of a single victim still occurs, there is evidence that the dynamics of this threat
have changed considerably over the last few years. Today the period of time between initial engagement with a child and an
offending outcome is often extremely short.
36. A key departure from traditional ideas of online grooming is that offenders focus on quickly gaining leverage over a victim rather
than first establishing a trusting relationship. A common feature of OCSE in 2012 was the investment of small amounts of time
by perpetrators in a large number of potential victims. This is a continuation of a trend identified in last year’s TACSEA. OCSE
is also sometimes referred to by commentators as ‘sextortion’. This term is not, however, considered appropriate to describe
serious and harmful criminal conduct.
37. The restrained influencing of a child over several months has been largely replaced by rapid escalation to threats, intimidation
and coercion. Both features are a symptom of the availability of thousands of potential victims online at any one time.
Reporting in 2012 highlighted the existence of multiple victims, sometimes in their hundreds, as a key feature of OCSE.
38. The objectives of OCSE have also evolved in recent years and such conduct can lead to a range of offending outcomes. Most
of these take place online, such as deceiving children into producing indecent images of themselves or engaging in sexual chat
or sexual activity over webcam. OCSE can also lead to offline offending, such as meetings between an adult and a child for
sexual purposes following online engagement.
Prevalence and risk factors
39. In 2012, CEOP received a total of 1,145 reports of OCSE behaviour from members of the public. While overall levels of public
reporting increased by almost 10%, the number of OCSE reports fell by over a quarter between 2011 and 2012.
40. More than two thirds (69%) of 2012 reports describe OCSE contact from an adult that failed - as a result of victim vigilance
- to result in full offending at the point of reporting. Increasing risk awareness by young people is evident from an analysis of
many of these reports. Such contact is, however, highly indicative of the early stages of OCSE and illustrates the ‘scatter gun’
approach to victim selection used by many offenders. It is therefore not possible to establish how many others have fallen
victim to these apparently unsuccessful approaches.
41. Many of the reports are made once the OCSE offending has begun to manifest itself but before it has escalated significantly. A single self-generated indecent image, for example, may have been exchanged in response to an approach by an offender.
In such circumstances, the victim often confides in a parent or other trusted adult resulting in a report to CEOP.
10
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 2: Online child sexual exploitation (OCSE)
42. Analysis reveals that 13 and 14 year olds represent the largest single victim group in reports to CEOP, accounting for 35%
of the total with 11 to 12 year olds accounting for 26% and 15 to 16 year olds accounting for 22%. Female victims were
apparent in 80% of all reports. This is illustrated in the table below:
TABLE 1: OCSE Victim demographics 7
Age range
% of OCSE reporting
in 20118
% of total OCSE reporting
in 2012 9
<10
14
8
11 to 12
27
26
13 to 14
36
35
15 to 16
23
22
> 16
0
5
Male
21
17
Female
79
80
43. In general, the initial offending outcome from OCSE takes place entirely online but it is recognised that such engagement can
in turn lead to offline meetings for sexual purposes. In a representative sample of OCSE reports received by CEOP in 2012,
offline meetings were apparent in 6.8% of cases. This is a reduction in prevalence from 11.7% in 2011.
44. In 2012, where it was possible to establish, 19% of public reports of OCSE involved the use of webcam. It is likely that the true
figure is substantially higher. The use of webcam has been identified as a key part of the grooming process and can be used to
reinforce, strengthen and maintain offending behaviour 10. An assessment of these reports reveals offenders using webcam to
show indecent imagery to victims in addition to capturing it.
45. Where it could be determined from the reports, the most common offending environment was social networking (SN) at
48.5%, with instant messaging and chat accounting for another 31%. OCSE on gaming sites and mobile phones featured in
a total of 10% of reports. Of interest is that 16% of reports received during 2012 involved multiple online environments.
46. In 2012 the average SN user in the 8 to 11 year old age range does not know 12% of their SN ‘friends’ in the real world. This
equates to around 11 ‘friends’ per child. These figures increase dramatically for 12 to 15 year olds where 25% of ‘friends’ (72
per child) are only known to them online. A quarter of 8 to 11 year olds and a third of 12 to 15 year olds communicate with
people online that they don’t know in the real world. These figures represent a 100% increase and 41% increase respectively
on previous year’s values.11
Self-generated indecent imagery (SGII)
47. The practice of self generating indecent imagery, a risk factor for OCSE, was highlighted in the previous TACSEA as becoming
increasingly widespread among children and young people in the UK. With the internet in 2012 increasingly defined by user
generated content, it was inevitable that CEOP would continue to receive reports of indecent images being self generated by
children. Analysis indicates that 21.2% of reports received in 2012 concerned the sharing of SGII. This figure has remained
relatively stable since 2011 where the figure was 22%.
48. Approximately 16,200 self generated still images were contained within these reports. From a total of 166 reports of moving
IIOC received over the same period (where it was possible to make an assessment) 113 (68%) appeared to be self generated.
The data refers to cases where age and gender could be determined. Of the cases reported in 2012, in 4% of cases age could not be determined and in 3% of cases gender was unknown.
Sample of 100 OCSE reports received from members of the public in 2011.
9
Sample of 350 OCSE reports received from members of the public 2012.
10
Webster, S., Davidson, J., Bifulco, A., Gottschalk, P., Caretti,V., et al. (2012) European Online Grooming Project : Final report – Executive Summary
11
DoubleClick Ad Planner (Google) U.S. Demographics.
7
8
11
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 2: Online child sexual exploitation (OCSE)
49. CEOP continues to assess the majority of this imagery as having been freely produced by young adolescents in the course
of developmentally appropriate behaviour not involving coercive or exploitative conduct by an adult. It is likely, however, that
a smaller number of these images will have been produced as a result of deception or coercion or will go on to be used for
such purposes. SGII reports are risk assessed by CEOP staff at point of receipt based on indicators within the imagery.
50. In order to better understand the image format, demographics and behaviours associated with SGII, a representative sample
of 120 still and all 113 moving SGII received by CEOP in 2012 was analysed. The results are shown in the table below:
TABLE 2: Analysis of Self Generated Indecent Imagery 12
Category
Still SGII
Moving SGII
Male subjects
18%
55%
Female subjects
82%
37%
Male and female subjects
0%
6%
Subject aged over 10
93%
51%
Subject aged under 10
0%
33%
White subjects
86%
83%
Non-white subjects
4%
10%
SC level 1
96%
61%
SC level 2
4%
27%
SC level 3
0%
1%
SC level 4
0%
11%
SC level 5
Nil
Nil
51. The prevalence of younger subjects in the moving image sample is of concern. A common theme within the sample, however,
particularly those featuring younger male children, is that playful rather than sexual activity is the norm. While recording
such material and sharing it online carries obvious risks to the young children concerned, the activity depicted often has the
appearance of normal developmental behaviour. In stark contrast, moving images also showed a greater tendency towards
more sexualised content than still imagery with, 10% depicting penetrative sexual activity.
52. Most SGII seen by CEOP during the year appears to be captured in domestic settings such as bedrooms and bathrooms.
In 2012, 43% of children use the internet in their bedroom with 24% of 8 to 11 year olds and 55% aged between 12 and
15 using the internet alone. Given the ease with which images can be taken and uploaded to the internet, the high level of
smart phone ownership amongst these age groups outlined in paragraph 54 is likely to contribute to this. The direct parental
supervision of children’s internet use is increasingly unfeasible.
12
12
Given the nature of the imagery it was not possible to identify relevant information in all cases.
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 2: Online child sexual exploitation (OCSE)
Summary
53. Whether and to what extent SGII prevalence is illustrative of the threat to children from OCSE remained unclear in 2012. Freely
generated imagery shared privately and consensually between adolescents remains unlikely to find its way into the hands of
adult offenders. Such imagery may find its way onto the open internet, however, where the original recipient subsequently
posts the imagery onto an open social networking profile or an image hosting platform, either through ignorance of the
consequences or a more malicious motive. It is also possible that a young person will ill-advisedly post a self generated image
of themselves onto a similar platform.
54. As predicted in last year’s TACSEA, smart phone ownership amongst children increased during 2012. Smart phone ownership
amongst 5 to 15 year olds increased from 20% to 28% during the year, driven by a 21% increase in ownership by the 12 to 15
age group to 62%13. It is now the preferred method of accessing the internet for the latter age range. With live video chat now
widely available on smart phones through applications such as Skype, Facetime and Tango, the threat from the use of these
facilities is likely to increase in 2013.
55. The effect of multiple time zones means that there are always high numbers of children online across the world. Reporting in
2012 highlighted the existence of multiple OCSE victims, sometimes in their hundreds. The nature of online communication
facilitates a ‘scatter gun’ approach where perpetrators are prepared to play the averages in their pursuit of victims.
Ofcom Children and Parents: Media Use and Attitudes October 2012.
13
13
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 3: Transnational child sexual abuse
56. Those UK nationals who sexually abuse or exploit children overseas are frequently denoted by the term ‘Travelling Sex
Offender’ (TSO). Although widely used across the global child protection community, the term may be misleading on two
counts. Firstly the use of the term infers travel as a key component of the offending. This is inaccurate as some of the most
prolific offenders often permanently reside overseas.
57. Secondly the term can appear to suggest that the threat is from Registered Sex Offenders (RSO) who travel. Historically this
has led to a disproportionate focus on the threat from RSOs who travel overseas. Reporting suggests that RSOs represent
a small minority of those UK nationals who commit sexual crimes against children overseas.
58. For the purposes of the TACSEA, the term ‘Transnational Child Sex Offender’, or TCSO, will be used to emphasise the specific
threat to children outside of the UK.
59. Research undertaken by CEOP suggests that overseas child sexual exploitation and abuse perpetrated by TCSOs can be
described using a framework of characteristics, including the offending type as described in the table below:
Transnational Offending Type
TRANSIENT
EMBEDDED
This describes overseas offending which is not the result
Embedded overseas offending is that which is repeatedly
of an extended grooming process. It may be facilitated
perpetrated against the same child or children as the
by the commercial sexual exploitation of children or by
result of an extended grooming process. This may
country conditions and vulnerabilities which facilitate the
include the grooming of significant others and the
14
transactional exploitation of children . Transient offending
community as well as the victim. Perpetrators are more
is generally perpetrated against a child or children for a
likely to be permanently or semi-permanently residing
short time, although it may be prolific. Some overseas
overseas, although not exclusively.
residents will perpetrate transient offending.
Individuals who embed themselves in local society to gain
Though the transient offending of an individual will not
access to and abuse children pose a threat to children
be perpetrated against the same child or children for a
that is distinct from that posed by transient TCSOs.
prolonged period, those exploited in the commercial sex
CEOP assesses the sexual abuse perpetrated by resident
industry are exposed to perpetual and prolonged abuse
TCSOs is likely to be prolonged, extensive and prolific.
by multiple abusers. A transient offender is perpetuating
Such abuse is associated with poorer outcomes for child
this continual abuse, although he or she may only be
victims15.
personally responsible for a small part of it.
60. It is important, however, not to label the offender as ‘transient’ or ‘embedded’ because offending behaviour should be seen
as a potential continuum, whereby an individual may begin offending transiently then move into embedded offending, or
vice-versa.
TCSO offending in 2012
61. A total of 104 UK nationals convicted or suspected of committing sexual crimes against children overseas were reported to
CEOP in 201216. Following arrests overseas for child sex offences, 79 individuals sought British Embassy consular assistance.
This is a notable increase from the 2011 total of 66, which is largely attributable to a rise in requests for consular assistance in
the USA and Canada.
Where a child receives a non-cash commodity in return for sexual activity, for example food or clothing.
Mian, M., Marton, P. & LeBaron, D. ‘The Effects of Sexual Abuse on 3-5 Year Old Girls’, Child Abuse & Neglect, 8 (1996), 731-745, and Kendall-Tackett, K.A., Williams, L.M & Finkelhor, D.
‘Impact of Sexual Abuse on Children: A Review & Synthesis of Recent Empirical Studies’, Psychological Bulletin, 113 (1993), 164-180.
16
Comprising 64 convictions, 20 arrests, 14 suspects, and 5 where this information is not known. Data source CEOP intelligence and overseas conviction data provided by ACPO Criminal
Records Office (ACRO).
14
14
15
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 3: Transnational child sexual abuse
62. Of those 104 reported to CEOP, the number who could be confirmed as previously unknown to police (i.e. not Registered Sex
Offenders) prior to their offending or suspected offending, overseas was five times larger than the number who were confirmed
as RSOs. Data from previous years further supports the inference that RSOs make up a minority of British transnational child
sex offenders. There is no evidence to suggest that this pattern will change so it is likely that most TCSOs will continue to be
previously unknown to UK law enforcement.
63. Almost all of those reported to CEOP as suspected or convicted of offending overseas in 2012 were male (99%), with the
majority (67%) being above the age of 40. This tallies with a study of 42 United States transnational child sex offenders which
found 87% of the sample to be over the age of 4017. Four individuals below the age of 20 were identified, whilst the most
common age range was 51-60. This is in contrast to the markedly younger age demographic found in samples of child abuse
image offenders18.
TCSO age ranges 2012
25
20
15
10
5
0
Below 20
21 - 30
31 - 40
41 - 50
51 - 60
60+
Unknown
Destination countries
64. Just under one third (31%) of the data collected in 2012 concerned countries in Asia, Africa, East and South Europe, the
Caribbean and Central America. In contrast, 68% of cases related to countries with advanced economies in North and West
Europe, North America and Australasia where socio-economic and cultural factors do not as obviously create a favourable
environment for offending. This is a similar proportional division of cases as seen by CEOP in 2011. To a degree this reflects
the effectiveness of the law enforcement response to CSEA in those countries, as well as information sharing and notification
procedures.
65. Analysis of the travel habits of the general British populace shows that, without exception, countries that attract peak numbers
of visits from RSOs also attract peak numbers of visits from the British public. When general public travel data is overlaid with
RSO travel notifications, some countries stand out as locations to which proportionally more RSOs notify travel than the British
public travel or migrate to. These include Spain, Turkey, Egypt, Cyprus, Bulgaria, Pakistan, and to a lesser extent across four
Caribbean islands19, Bangladesh, Lithuania, Greece, Malta and the Gambia.
The Protection Project, International Child Sex Tourism: Scope of the Problem & Comparative Case Studies, (The John Hopkins University, 2007), p.39.
CEOP, A Picture of Abuse, (CEOP, 2012).
19
Antigua & Barbuda, Barbados, Dominican Republic and Saint Kitts and Nevis.
17
18
15
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 3: Transnational child sexual abuse
66. It is difficult to draw meaningful conclusions from this data due to the fact that the percentage differences are small and the
public travel data is taken from a sample survey 20 and is therefore an informed estimation. More targeted in-depth country
research is required to determine if this data can be used to assess the risk posed to children by these offenders.
67. Colombia, Venezuela, Vietnam, Japan, Russia and Thailand are identified on hidden internet forums in 2012 as presenting
favourable circumstances for sexually abusing children. Whilst these forums are not necessarily created by (nor their
membership restricted to) UK nationals, they may give an indication of some potential new geographical threat areas not yet
specifically identified in hard data.
68. South East Asia continues to be an area targeted by British TCSOs perpetrating both transient and embedded offending. Law
enforcement in the Philippines, Thailand and to a lesser degree Cambodia has become increasingly robust in dealing with child
sexual exploitation and abuse, to the extent that prostituted children reportedly are no longer visible on the streets in some of
the main sex tourism destination towns in Thailand. This appears to be borne out by reports from 2012 concerning Thailand
which have been notably lower in 2011 and 2012 than previous years. This is likely to cause displacement to other countries
both in the region – such as Laos and Vietnam - and further afield. A recent survey of 361 international tourists to South East
Asia found that 20% of those surveyed had witnessed children working in the sex industry.
69. Information from internet chat forums supports the displacement hypothesis and indicates that the threat from transnational
British offenders may be reducing to an extent in Thailand as well as in India and the Philippines. Chat within these forums
suggests that organised commercial child sexual exploitation may have become more difficult for offenders to identify in
these countries.
70. Whilst hidden internet chat forums indicate that opportunities for offending may be diminishing in India, CEOP received
increased information relating to embedded TCSO offending in Bangladesh. Case data suggests that British TCSOs are
exploiting the vulnerability of street children in the country – of which there are estimated to be 500,000. A 2006 21 report
found that young boys living on the street, migrant boys and boys working as child labourers are especially vulnerable to
sexual exploitation.
71. There are indications that TCSO activity in South America may increase in the next few years. The Association of British Travel
Agents marked Colombia in particular as a UK tourist destination ‘to watch’ for 2013. Furthermore, Brazil is also likely to
receive an increase in UK tourists over the coming years as a result of its burgeoning economy and hosting of the FIFA World
Cup and the Olympics Games in 2014 and 2016 respectively. It is expected that both the adult and child sex markets will
expand during this time, leading to an increased risk of offending by UK nationals. Although reliable comparative data is hard
to come by, commercial sexual exploitation of children in Brazil is currently thought to be the most prevalent in the Western
hemisphere, possibly second only to Thailand globally.
72. Internet chat forums recommend countries that are severely deprived or that have suffered environmental disaster as
particularly advantageous for those wishing to sexually exploit children. It is likely that offenders would seek to exploit children
in countries where social, environmental or even political upheaval has resulted in displacement of children from their homes
and families, who would ordinarily offer them protection. The chance that local authorities will be in a position to identify and
disrupt or prosecute child sex abusers in these countries is particularly low.
20
16
21
Of 800,000 people.
ECPAT International and INCIDIN Bangladesh. Situational Analysis Report on Prostitution of Boys in Bangladesh, June 2006.
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 3: Transnational child sexual abuse
Summary
73. If we accept the assumption that there are significant gaps in the amount of hard data we have in relation to countries with
lower law enforcement capability and greater child vulnerability we also acknowledge that the prevalence of British TCSO
offending in these countries is likely to be higher than that reported. This is partially evidenced by the comparatively large
amount of hard data there is concerning Thailand and Cambodia where concerted law enforcement and NGO efforts to
combat CSEA are showing results and where CEOP has strong partner links.
74. CEOP analysis suggests that where similar CSEA facilitating factors exist, offending will be uncovered through proactive
investigation alongside partner relationship and capacity building in-country. Arrests in both hemispheres demonstrate that no
country or continent can be ruled out as a destination for transnational child sex offenders.
75. The International Child Protection Certificate (ICPC) developed by CEOP and the ACPO Criminal Records Office was launched
in October 2012. The ICPC provides a criminal records check for UK nationals who may not be currently resident in the
UK. The aim of the initiative is to prevent individuals with prior convictions for child sex offences in the UK from working with
children abroad. To date there has been a positive uptake with a total of 2,567 certificates applied for.
76. Initiatives such as the ICPC will doubtless have a positive impact on the threat from TCSOs, who seek to exploit children by
gaining a position of trust overseas and thereby perpetrate embedded overseas offending. Mitigating the overall threat from
TCSO offending, however, remains particularly problematic. The inevitable cultural, legal, linguistic and political barriers hamper
efforts by the UK child protection community to counter this threat.
77. CEOP’s International Child Protection Network (ICPN) carried out partner relationship and capability building activities in
Thailand, the Philippines, Vietnam, Cambodia, Ukraine and Spain in 2012. Intelligence continues to support activity in
these regions. It is clear that similar activity is required in other countries identified in this document to increase the UK’s
understanding of offending across a broader geography, improve information sharing and share knowledge in relation to the
international investigation of child sexual abuse and exploitation committed by British TCSOs.
78. Under the G8 Presidency of the Roma-Lyon Group and a joint US/UK initiative, CEOP will host an international symposium on
‘Transnational Child Sex Offenders’ from the 7th to the 9th October 2013. The objectives are to discuss resident, transient
and virtual offenders, with the intention of further increasing international co-operation against this key threat.
17
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 4: Contact child sexual abuse
79. The focus of this threat is on contact sexual offending against children by non-related adults. In the interests of precision,
the term contact sexual abuse (CSA) is used in preference to ‘child sexual exploitation’ (CSE) which, whilst in common use,
can be given a much wider meaning encompassing certain forms of non-contact sexual offending. For the purposes of this
assessment, CSA involves lone and group offenders and sexual offending associated with street gang culture.
80. There are a number of reasons why it is difficult to assess precisely the extent of contact sexual abuse by reference to
available data. Firstly, more than in most other types of offending, it is likely that contact sexual abuse is significantly underreported by its child victims. This may be due to the fact that children and young people do not always readily identify
themselves as victims due to the nature of the grooming process. Also, there may be a reluctance of child victims from some
cultures to come forward, which may lead to a disproportionate number of victims of white ethnicity in samples of recorded
CSA. Variations also exist in the manner in which CSA is recorded within the criminal justice system. Offenders face a range
of potential charges including rape, sexual assault and trafficking, which may not all be recorded as having an underlying
CSA context.
81. Terms such as ‘grooming’, ‘street grooming’, ‘localised grooming’, ‘trafficking’ and ‘group associated child sexual exploitation’
are often used interchangeably, despite relating to very specific typologies or behaviours. This can lead to unacknowledged
disparities between data sets collected from different sources.
82. With these limitations in mind, CEOP undertook a data collection exercise encompassing the 43 Police Forces of England and
Wales. Details of contact child sexual offending known or suspected to have taken place in 2012 were requested. A total of
31 forces responded, although not all were able to provide the full range of data as a result of different recording categories,
methodologies and standards. Of these, 25 forces identified a total of 2,120 lone perpetrators involved in either suspected or
confirmed cases of non-familial contact child sexual abuse in 2012. In comparison, all 31 forces reported a cumulative total of
65 group and gang associated offences.
Contact child sexual abuse by lone offenders
83. The sample obtained from the data collection illustrates that in 2012, whilst under-reporting must be taken into account, UK
police forces became aware of a significantly greater number of incidents of non-familial contact sexual abuse involving lone
offenders than involving groups or street gangs.
84. On-street interaction was found to be one of the most frequently recorded methods of initial contact between such offenders
and their victims, with introductions by third parties being recorded. These methods of victim engagement are also commonly
seen in group offending suggesting that some perpetrators, while ostensibly offending alone, may in fact have group affiliations.
Contact child sexual abuse by group offenders and offending associated with street gangs
85. Whilst often used as a single typology, group offending and offending associated with street gangs, appear to share few
characteristics. Public awareness of offending by groups increased considerably in 2012 following a small number of high
profile cases which attracted considerable public commentary. The extent to which these cases are representative of the wider
threat from group offending is, however, unclear.
86. The most comprehensive prevalence study to date of group offending and gang associated offending was undertaken in 2012
by the Office of the Children’s Commissioner (OCC) 22. In its interim report, the OCC identified 2,409 children as victims during
a 14 month period from August 2010 to October 2011 at the hands of 1,514 offenders. A broader definition was used by the
OCC in its request for data which may mean the study is not directly comparable to that undertaken by CEOP for this document.
22
18
“I thought I was the only one. The only one in the world”, The Office of the Children’s Commissioner’s Inquiry into Child Sexual Exploitation in Gangs and Groups, Interim Report,
November 2012.
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 4: Contact child sexual abuse
87. During 2012, attempts at defining group offending have led to the emergence of what is seen as a single typology. However,
analysis of available data suggests that group offending may be less homogeneous and its parameters less well defined than
a single label suggests. Two typologies emerge from CEOP analysis:
Type 1: Group offending targeting victim vulnerability
The focus here appears to be on the sexual abuse of teenagers and young adults on the basis of their vulnerability, rather
than as a result of a specific preferential sexual interest in children. Members of the offending group can be linked to
one another in a number of ways. Many have loose associations, others present as more formal networks or organised
groups with criminal or business ties rather than friendship. The abuse is carried out by more than one perpetrator and
can involve individual or multiple victims. CEOP assesses that type 1 offenders are unlikely to identify themselves as
having a sexual interest in children, but molest children because they are vulnerable to sexual exploitation 23.
Type 2: Group offending as a result of a specific sexual interest in children
In contrast, perpetrators of type 2 group offending have a similar demographic make up to that of indecent image
offenders as described in the CEOP 2012 Thematic Assessment ‘A Picture of Abuse’. They present as having a longstanding sexual interest in children with some having a synergy with what has been described as a paedophile ‘ring’.
Whilst a small number of cases of this type have been noted during 2012, real-world networking between those with
a sexual interest in children appears uncommon.
88. The definitions identified above formed the basis of CEOP’s police data collection exercise. The table below shows the results
of this data collection from the 31 ACPO forces that responded, broken down by group offending typology.
Type 1 group
Type 2 group
Gang associated
57
7
1
Type 1 group offending
From the data submitted by the 31 forces the following findings emerge:
Perpetrators
89. Group size varied considerably within the sample, ranging from pairs of offenders to groups comprising 25 individuals. Small
groups of four or less were the most prevalent, accounting for 67% of groups. Large groups of offenders did not, however,
automatically equate to large numbers of victims.
90. The most common offender age range was between 20 and 30, accounting for 53% of the offending population. This was
closely followed by those aged under 20, which accounted for a further 24% of offenders.
91. Ethnicity descriptors remain imprecise and as not all police forces responded to CEOP’s information request, the data is
incomplete. Using broad groupings, all ethnicities were represented in the sample, however, a disproportionate number of
offenders were reported as Asian. Of the 52 groups for which usable ethnicity data was provided, 26 (50%) comprised all
Asian offenders, 11 (21%) all white offenders, 9 (17%) groups comprised offenders from multiple ethnicities, 4 (8%) comprised
all black offenders and there were 2 (4%) groups of exclusively Arab offenders. Of the 306 offenders whose ethnicity was
provided for type 1 offending, a total of 75% were categorised as Asian, 17% were categorised as white, and the remaining
8% were categorised as black (5%) or Arab (3%). This is in contrast to those identified in type 2 group offending, who were
reported as exclusively of white ethnicity.
Neil Malamuth and Mark Huppin, Drawing a Line on Virtual Child Pornography: Bringing the Law in Line with the Research Evidence, 31 N.Y.U. REV. L & SOC. CHANGE 773, 823 (2007).
23
19
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 4: Contact child sexual abuse
Victims
92. From the 57 type 1 cases reported to CEOP, 144 victims were identified. Victim demographics yielded similar results to that
of other studies, including the Office of the Children’s Commissioner's interim report mentioned earlier, although not with the
same identified or suspected victim prevalence. Gender was provided for 118 victims, all of which were female.
93. The most prominent victim age group identified was children aged between 14 and 15. This accounted for 57% of the sample,
followed by 16 to 17 year olds which accounted for a further 40%. The remaining 3% of victims were aged between 12 and 13.
94. Of the 144 victims reported, 100 were described as having at least one identifiable vulnerability. These included the
consumption of alcohol and drugs, mental health issues, regular missing episodes, living in local authority care or being known
to the police and/or social services. Over half of the victims reported were in local authority care.
95. Those victims who were not reported as vulnerable were often associated with vulnerable children through friendship. However
it is difficult to assess whether this sample of looked after children is representative or whether care homes are simply better
equipped to identify the offending.
96. In 97% of cases reported to CEOP, victims of type 1 offending were categorised as white. The comparative levels of freedom
that white British children enjoy in comparison to some other ethnicities may make them more vulnerable to exploitation 24.
They may also be more likely to report abuse. This is an area requiring better data and further research.
Type 2 groups
Perpetrators
97. Of the six 25 type 2 offending scenarios where group size was reported, half involved pairs of offenders. Group sizes ranged
from two to five.
98. Offenders were all white and 18 of the 19 offenders were male. The one female was offending in conjunction with a male offender.
99. Offenders were generally older than those identified under type 1 and groups that included younger members generally
included a significantly older offender.
100. Of the 12 offenders for which age data was provided; 33% of offenders were aged 51-60; 25%, 41-50; 17%, 31-40; 17%,
20-30; and 8% under 20. One possible explanation for a prevalence of older offenders in this sample may be the nature of
the offending. Type 2 offending appears to be an extension of lone offending, where the group is not an intrinsic part of the
offence, but is used as a means of extending the ability to access a victim.
101. A limited amount of data in relation to the criminal history of the individual offenders was reported. However of the three groups
where information was provided, one group involved a registered sex offender with an index offence of possession of indecent
images. Both members of this group were known to network with other offenders with significant criminal histories. The two
other groups involved people who had been suspected of child sexual offences, but had never been charged. In a separate
case, three of the five offenders had occupations that allowed them access to children and young people. These cases,
coupled with the age of the victims discussed below, indicate that type 2 offenders may show a greater likelihood of having
a specific sexual interest in children.
24
20
25
Reist, T.M. (2008) Getting Real: Challenging the Sexualisation of Girls.
One case reported focused on a group where the true number of participants was unknown. This case has been excluded when analysing group size.
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Key threat 4: Contact child sexual abuse
Victims
102. Victims of type 2 offending were generally younger than those identified in type 1 cases. Victim age ranged from 7 to 15, with
70% of victims being under the age of 13. A total of 8 out of 10 victims were male.
103. In comparison to the victims in type 1 offending, the majority of these children were living with their parents within the family
home. However, in two of the seven cases a lack of supervision of the child victim by parents was noted in reports.
Child sexual abuse associated with street gangs
104. Gang associated child sexual abuse is believed to take place where offenders see themselves and are seen by others, as
affiliates of a named criminal group with distinctive beliefs, attitudes and behaviours. Girls are often seen as sexual accessories
by young men in gangs and can be passed around group members as a rite of passage or to repay debts. This includes girls
having to perform sexual acts on gang members, or girlfriends of gang members being targeted and sexually and violently
assaulted by rival gang members. Rape and sexual assault by individual gang members and by the whole group is relatively
common 26. External research has shown that the victim and offender are generally known to each other prior to the offending
taking place, which is not normally the case in group offending.
105. Research carried out by the University of Bedfordshire 27 identifies that gang associated CSA is significantly under-reported.
This may be because victims or witnesses feel a kinship to other gang members, fear reprisal or feel that the authorities will not
believe or help them. The university is currently conducting research and fieldwork into gang associated sexual exploitation and
sexual violence in six different areas of England. Preliminary analysis from interviews identifies that perpetrators of gang sexual
violence are predominantly male, whilst victims are predominantly female.
106. In CEOP’s most recent data collection exercise detailed above, only one case of gang associated abuse was reported. The
sample provided is not sufficient to draw conclusions in relation to gang demographics.
Summary
107. A range of trends in relation to offender and victim demographics were identified in the data collected by CEOP from UK
forces. However, due to the small sample sizes, more research needs to be done before any firm conclusions can be drawn.
This along with ACPO plans to improve data recording and collection, will help to enhance the national picture. There would
be benefit in moving away from a single all encompassing term for such a wide variety of offending. This would ensure
a more consistent approach to data and intelligence collection and enable greater understanding of offender and victim
demographics. An increased understanding of offender behaviour and typologies, victim vulnerabilities and prevalence would
also facilitate more intelligent targeting of prevention resources.
108. Data suggests that in 2012, offending by lone perpetrators is significantly more prevalent than offending by groups or offending
by those associated with street gangs. Within group offending, small groups or pairs appear to be the most common group
size, with no correlation between the size or number of the group and the number of victims targeted. The extent to which a
pair of offenders can properly be considered group offending is questionable. Until the current confusion around terminology is
resolved by clear taxonomy, it is unwise to place too much emphasis on arbitrary categorisation.
Centre for Social Justice, 2009, Dying to Belong, An In-Depth Review of Street Gangs in Britain.
University of Bedfordshire, Research into gang-associated sexual exploitation and sexual violence, Interim report, November 2012.
26
27
21
Threat Assessment of Child Sexual Exploitation and Abuse 2013
Appendix A
Image Level
22
Image Description
1
Nudity or erotic posing with no sexual activity
2
Sexual activity between children, or solo masturbation by a child
3
Non-penetrative sexual activity between adult(s) and child(ren)
4
Penetrative sexual activity between child(ren) and adult(s)
5
Sadism or bestiality
Child Exploitation and Online Protection Centre
33 Vauxhall Bridge Road
London
SW1V 2WG
www.ceop.police.uk